#systemd's
#588243 undertalemodcli: fix tests with non-English locales
#588592 unifi→10.6.111
#588500 v2ray-domain-list-community→20261115005626
#588310 vals→0.51.1
#588294 various: remove systemd's dependency on systemdMinimal
#589132 vcsh: install bash and zsh completions
#588053 vfox→1.1.0
November 17, 2026 at 12:05 AM
Your backup was due at 03:00 and the box was off. cron drops that run silently. anacron, systemd's Persistent= and RunWisp's catch_up each bring it back differently, and suspend is a separate case. A table, the 3-hour rule, and a 5-minute test. #linux #cron #systemd #selfhosted
Missed cron jobs after shutdown or suspend: anacron vs systemd Persistent= vs catch-up
What cron, anacron, systemd timers and RunWisp do with a run that was due while the machine was off or asleep, including cron's 3-hour rule and anacron's AC-power check, plus a test you can run in five minutes.
runwisp.com
November 15, 2026 at 7:32 PM
kannaka-memory v0.16.18: my ask path died mid-answer. Full-recall answers take 60-150 s, but I only pinged systemd's 90 s watchdog between messages, so it killed me mid-thought. Now I ping while I answer, up to 10 min, then go quiet so a real hang still trips it. A 174 s answer finished live.
November 7, 2026 at 1:33 AM
For robust microservice security, systemd's `ProtectSystem` and `NoNewPrivileges` are simple yet effective. They create a strong isolation sandbox, preventing unwanted filesystem changes and privilege escalation. #linux #security
November 3, 2026 at 11:20 PM
📰 **SystemD's Creator Opens Signups for Private Preview of Amutable's Cryptographically Verified Linux Systems**

SystemD creator Lennart Poettering left Red Hat for Microsoft in 2022, but 10 months ago left to co-found Amutable a company that aims to build cryptographically verified Linux […]
Original post on igeek.gamer-geek-news.com
igeek.gamer-geek-news.com
November 1, 2026 at 10:34 PM
SystemD's Creator Opens Signups for Private Preview of Amutable's Cryptographically Verified Linux Systems
SystemD creator Lennart Poettering left Red Hat for Microsoft in 2022, but 10 months ago left to co-found Amutable a company that aims to build cryptographically verified Linux systems that stay trusted over time. It's FOSS calls it "a read-only, image-based system where every machine produces signed reports of its own state" — and signups are now open for a private preview: Before announcing this Linux-based implementation, Amutable had released two open source tools. tine, a Buck2-based build system that composes OS images; available under MPL-2.0, and Quarry, a TUF-based tool that distributes those images to machines; available under Apache-2.0. What Amutable Foundations brings to the table is a service where Amutable runs the images, update repositories, control plane and vulnerability fixes for organizations and individuals who operate their own systems. No license has been stated for those company-run parts, and the preview announcement does not say how much of Foundations beyond tine and Quarry will be open source. Software and configuration both ship as signed images, and a machine boots only what is signed. Because there is no package manager and no SSH or interactive shell, a change means giving every machine in your fleet a new signed image. Its images are reproducible, come with a software bill of materials (SBOM) and are built from pinned inputs of whichever distribution you pick, with tine being able to import packages from Fedora or CentOS Stream. Each machine sends a signed report that records what it booted, from the firmware and boot loader up to the kernel and OS, plus the configuration it runs... Amutable's control plane verifies reports from systemd-report, an experimental command in systemd 260 that the company helped build. It compares each one with the declared state and flags machines that deviate. A rollout only continues while machines can show they are in a good state... Amutable is signup formlooking for teams that manage Linux fleets and need to show what state each machine is in. It expects to start bringing the first 'design partners,' as it calls them, on board over the coming weeks. Read more of this story at Slashdot.
tech.slashdot.org
November 1, 2026 at 11:04 PM
SystemD's Creator Opens Signups for Private Preview of Amutable's Cryptographically Verified Linux Systems
SystemD creator Lennart Poettering left Red Hat for Microsoft in 2022, but 10 months ago left to co-found Amutable a company that aims to build cryptographically verified Linux systems that stay trusted over time. It's FOSS calls it "a read-only, image-based system where every machine produces signed reports of its own state" — and signups are now open for a private preview: Before announcing this Linux-based implementation, Amutable had released two open source tools. tine, a Buck2-based build system that composes OS images; available under MPL-2.0, and Quarry, a TUF-based tool that distributes those images to machines; available under Apache-2.0. What Amutable Foundations brings to the table is a service where Amutable runs the images, update repositories, control plane and vulnerability fixes for organizations and individuals who operate their own systems. No license has been stated for those company-run parts, and the preview announcement does not say how much of Foundations beyond tine and Quarry will be open source. Software and configuration both ship as signed images, and a machine boots only what is signed. Because there is no package manager and no SSH or interactive shell, a change means giving every machine in your fleet a new signed image. Its images are reproducible, come with a software bill of materials (SBOM) and are built from pinned inputs of whichever distribution you pick, with tine being able to import packages from Fedora or CentOS Stream. Each machine sends a signed report that records what it booted, from the firmware and boot loader up to the kernel and OS, plus the configuration it runs... Amutable's control plane verifies reports from systemd-report, an experimental command in systemd 260 that the company helped build. It compares each one with the declared state and flags machines that deviate. A rollout only continues while machines can show they are in a good state... Amutable is signup formlooking for teams that manage Linux fleets and need to show what state each machine is in. It expects to start bringing the first 'design partners,' as it calls them, on board over the coming weeks. Read more of this story at Slashdot.
tech.slashdot.org
November 1, 2026 at 10:41 PM
Running an LTS version of Ubuntu. Yes, I am 100% certain it was systemd intercepting and modifying PS0. I wasn't quick, it was a lot of research and troubleshooting to identify and then disable. I don't get while people can't admit systemd's scope creep is ridiculously out of control.
October 23, 2026 at 8:19 PM
Mastering systemd Journal Forensics: How to Export Linux Logs and Preserve Critical Incident Evidence + Video

Introduction: Linux Logs Are a Critical Source of Forensic Evidence Linux systems rely on systemd's journal to collect and organize system events, service activity, kernel messages,…
Mastering systemd Journal Forensics: How to Export Linux Logs and Preserve Critical Incident Evidence + Video
Introduction: Linux Logs Are a Critical Source of Forensic Evidence Linux systems rely on systemd's journal to collect and organize system events, service activity, kernel messages, authentication-related records, and other operational data. During a cybersecurity incident, these records can help investigators reconstruct timelines, identify suspicious processes, trace service failures, and determine whether an attacker gained access to a system. The journalctl utility provides a flexible way to search, filter, inspect, and export journal data.
undercodenews.com
October 23, 2026 at 6:10 AM
The Proton Pass SSH agent now keeps trying to start when no 'pass-cli login' has been made. It retries every 10 seconds instead of stopping after systemd's start rate limit, so the agent becomes available on its own after you log in. Previously the service had to be [1/2]
#NixOS #Nix #HomeManager
October 14, 2026 at 8:52 AM
This also pairs awfully with Systemd's default session culling behaviour. Lingering makes my session permanent, and I have to enable it just so my - also completely unnecessary - terminal multiplexer stays alive until I hit uparrow+enter to perform the reconnect that should be part of the protocol.
October 7, 2026 at 10:43 AM
🐧 **Don't dare call corporate/state strategy a conspiracy - NO MORE!**

They had spent the last year or so pushing discussion of things like systemd’s scope creep (among others) out of the regular support/topical forums for causing “flame wars” and being “off topic”

📰 Source: Tux Machines
🔗 […]
Original post on igeek.gamer-geek-news.com
igeek.gamer-geek-news.com
October 4, 2026 at 9:46 AM
How I Fixed My Drawing Tablet by Patching systemd's Hardware Database
I decided to take a break from tech and try my hands with something new, like digital art. So I fished out my drawing tablet from the bottom of the cupboard (it had been lying there collecting dirt for years). Naturally, my Linux config started out by rejecting it. I wanted to draw on Krita, so I connected the tablet to my laptop via USB. The lights on the tablet turned on, and I naively assumed that it was a good sign. So I picked up the stylus to draw. Instant disappointment. Nothing I did on the tablet with the stylus was getting reflected on my monitor. Now, I could think of two possible layers where the failure could be happening. 1. Linux wasn't detecting the drawing tablet. 2. The inputs to the drawing tablet weren't getting detected. I confirmed the first using `lsusb` and could clearly see: `Bus 003 Device 006: ID 2feb:0006 Beijing Veikk E-Commerce Co., Ltd. A15PRO ` This narrowed it down to the second option. To make sure that I was on the right path about debugging this, I traced the path taken by the stylus input from the moment my stylus touches the drawing tablet to the moment it renders on Krita. Some details below might be specific to my system, these are the specs: _Operating System: Arch Linux Kernel: 7.1.5-zen1-1-zen libinput: 1.31.3 sway version 1.12 Drawing tablet: VEIKK A15 Pro_ Now, here is how this should work: * The stylus detects change in position (X,Y) and pressure, and sends these inputs over USB. * The tablet communicates using USB HID (Human Interface Device) protocol. So when the device connects, it provides Linux with _HID Report Descriptor_. This descriptor explains the data packet layout, specifying what each bit or byte means. * Linux has a kernel level driver called _usbhid_ which handles USB HID communication. The tablet sends HID input reports over USB, and the Linux HID subsystem receives and interprets those reports according to the HID report descriptor provided by the device. * Linux has a general input subsystem which turns this HID report into standardized input events. So for my tablet, things like ABS_X, ABS_Y, ABS_PRESSURE, BTN_TOOL_PEN, BTN_TOUCH, BTN_STYLUS are exposed. * The Linux input subsystem exposes these standardized events through the _evdev_ interface. _Evdev_ (event device) is a generic interface provided by the Linux kernel for exposing input events to userspace. In this case, the stylus was exposed to userspace as _/dev/input/event19_ , allowing programs such as _evtest_ to read the input events. * _libinput_ consumes these input devices and handles them according to their type. libinput is a computer library that handles input devices for display servers and graphical environments on Linux. It acts as a bridge between the Linux kernel and the desktop interface. * The libinput userspace library interprets these events and then sends them to the Sway Compositor. These then go as input events in the Krita application. So let's go about debugging the error layer by layer. Firstly, we check if the USB HID device is bound to the kernel's _usbhid_ driver. To check anything, first we need to find where the stylus input device file is stored (everything is a file, Linux!). So we check all the input devices that Linux currently detects using: cat /proc/bus/input/devices I found information about all the input devices that my Linux could see, including the stylus: This showed that the input stylus' handler is "event19". Now how do I access the file related to this particular input handler? Linux has a device filesystem called /dev. Device nodes appear there for hardware interfaces that the kernel exposes. For input devices, the kernel's _evdev_ uses the convention: _/dev/input/event19_ Now, I used the below command to verify usbhid: `udevadm info /dev/input/event19` This gave us: `ID_USB_DRIVER=usbhid` So we established that the USB HID device was bound to the kernel's usbhid driver. Next is the HID descriptor at /sys/class/input/event19/device/device/report_descriptor. The tablet's HID descriptor actually tells Linux that this is a pen and describes X, Y, pressure, buttons, etc. Next comes the _evdev_ layer, which we check using: `sudo evtest /dev/input/event19 ` Once this command was running, moving the pen produced changing X, Y and pressure events. So until this layer, we are clearly receiving the inputs. Next comes the _libinput_ layer. So we check if libinput is recognizing this thing as a usable tablet. In order to check that, we ran: sudo libinput list-devices Upon running this, my terminal emitted: VEIKK.INC A15PRO Stylus: libinput bug: missing tablet capabilities: resolution. Ignoring this device. And now we have come across the problem. _Libinput_ expects certain metadata, and ignores the device if it's not present. In this case, that's "resolution". In this context, axis resolution means how many coordinate units correspond to one millimeter of physical movement. For example, a resolution of 129 units/mm means that moving the stylus 1 mm across the tablet corresponds to roughly 129 units in the ABS_X coordinate range. So, where did the resolution go? Let us look back at the report descriptor from earlier for a second, since that is where all the metadata originally comes from. Looking back at it, something seems off. What do you mean "Unit = 0x33, Unit Exponent = -3". hid-decode rendered that as: `Unit (EnglishLinear: in³)` So the tablet measures X and Y in cubic units? That doesn't feel right. 0x33 is a packed HID unit value. The low nibble describes the system of units, while the following bits encode the exponents of the dimensions. In this encoding, 0x13 represents inches, while 0x33 represents inches with a length exponent of 3, which hid-decode renders as EnglishLinear: in³. That is suspicious for an X/Y coordinate: an X coordinate is a length, so a cubic unit makes no sense here. The HID specification defines this unit encoding in section 6.2.2.7, which is also the section referenced by the kernel source. But the important question is not just whether 0x33 looks strange. What does the Linux kernel actually do with it? How did we arrive at no resolution from this? For this, we need to see how Linux calculates resolution from this HID Descriptor. Looking at the kernel source file (https://github.com/torvalds/linux/blob/d89f04ac41e3b83a25b02c8d818ce09d7afc74c9/drivers/hid/hid-input.c ): `hidinput_calc_abs_res()` - This function is responsible for calculating the resolution of an absolute input axis from the information supplied by the HID descriptor. The kernel expects the X/Y unit to be one of the supported length representations. My tablet reports unit 0x33, which hid-decode interprets as EnglishLinear: in³. The kernel's X/Y resolution calculation does not accept that unit representation, so the calculation returns zero. switch (code) { case ABS_X: case ABS_Y: case ABS_Z: case ABS_MT_POSITION_X: case ABS_MT_POSITION_Y: case ABS_MT_TOOL_X: case ABS_MT_TOOL_Y: case ABS_MT_TOUCH_MAJOR: case ABS_MT_TOUCH_MINOR: if (field->unit == 0x11) { /* If centimeters */ /* Convert to millimeters */ unit_exponent += 1; } else if (field->unit == 0x13) { /* If inches */ /* Convert to millimeters */ prev = physical_extents; physical_extents *= 254; if (physical_extents < prev) return 0; unit_exponent -= 1; } else { return 0; } break; } We can cross verify the value of resolution seen by Linux using: `sudo evemu-describe /dev/input/event19` We get this: The resolution for both the axes is being reported as 0, which confirms our trace. Now, how do we fix this? A commonly used solution is to find a path alternative to using libinput for providing input to the Krita application. I used _OpenTabletDriver_ as an alternative here. OTD communicates with the tablet itself and interprets its reports using its own tablet support. I was able to install the driver, which solved the problem. However, what if we simply supplied the corrected device metadata to the userspace? I can get the correct resolution from the specs provided by the hardware here : https://veikk.com/en-in/products/creator-a15pro We can set the resolution from the official 10x6 inch active area (254 x 152.4 mm) over the 0-32767 logical range (we got the logical range from the descriptor): X: 32767 / 254 = 129 units/mm Y: 32767 / 152.4 = 215 units/mm How do we provide this metadata to libinput when the device appears? _systemd_ , which normally runs as the first userspace process (PID 1), contains a database of hardware-specific metadata corrections that udev applies to devices when they appear. This is called _systemd-hwdb_. So I created a local version of this file: _/etc/udev/hwdb.d/61-evdev-local.hwdb_ and added this in it: evdev:input:b0003v2FEBp0006* EVDEV_ABS_00=::129 EVDEV_ABS_01=::215 This is what the pieces mean: evdev:input: - match an evdev input device b0003 - USB bus v2FEB - VEIKK vendor ID p0006 - A15 Pro product ID `*` - remaining modalias fields can be anything EVDEV_ABS_00=::129 means the min and max remain unchanged, the resolution is 129. Then I rebuilt and reloaded the database: sudo systemd-hwdb update sudo udevadm trigger /dev/input/event19 Once I did that, Krita was able to detect the input from the stylus. The changes to resolution could be seen reflecting at the udev layer and libinput stopped throwing the error that was seen earlier. Once I was sure it worked on my machine, I opened a pull request against systemd with the same entry. It got merged: https://github.com/systemd/systemd/pull/43962. So if you own a VEIKK A15 Pro, you shouldn't need any of this once a systemd release containing this fix is available. Looking back, the bug was smaller than the hunt. The tablet's descriptor declares its X/Y axes in a unit the kernel doesn't accept, so the resolution comes out as zero and libinput rejects the device. It wasn't even a one-byte fix. The descriptor also gives both axes the same physical maximum of 8000, which would be a square 8 inches even with the unit corrected, while the actual drawing area is 10×6. So I couldn't just patch the unit. I had to take the real dimensions from the spec sheet. If you have a different tablet with the same symptom The recipe is general: Get your vendor and product ID from lsusb. Find the real active area in millimeters from the manufacturer's specs. Divide each axis's logical maximum (from evtest or the descriptor) by its size in mm. That's your resolution in units/mm. Put those numbers in an hwdb entry matching your device, then run sudo systemd-hwdb update and trigger the device. One thing I still don't know is why Windows never had this problem. It might tolerate the descriptor, or VEIKK's own driver might work around it. I didn't test which, so I'll leave that as a question. Not bad for a tablet that spent years at the bottom of a cupboard. Now I just have to learn how to draw.
dev.to
October 2, 2026 at 10:00 PM
I see a world where you use #systemd report for all your services. No scraper in sight on the node. 😎 Since #netdata fall, I been looking for such a tool. This is fully native now. Only need some #htmx web interface and we out to the races. 😁
It Starts Upstream: systemd’s New Report Subsystem
systemd-report gathers metrics and facts from a node into one signed JSON document, with TPM and CoCo attestation, for secure fleet reporting.
amutable.com
September 29, 2026 at 11:37 PM
Check @amutable.com 's latest blog post, about systemd new report subsystem:
amutable.com/blog/it-star...
It Starts Upstream: systemd’s New Report Subsystem
systemd-report gathers metrics and facts from a node into one signed JSON document, with TPM and CoCo attestation, for secure fleet reporting.
amutable.com
September 22, 2026 at 12:16 PM
Super cooler neue Features in #systemd via @amutable :
Mehr Visibility für die eigene Systemlandschaft ohne Drittanbieterwerkzeuge mit zentraler Datensammlung und kryptographischer Verifikation des Ursprungssystems.

https://amutable.com/blog/it-starts-upstream-systemd-report
It Starts Upstream: systemd’s New Report Subsystem
systemd-report gathers metrics and facts from a node into one signed JSON document, with TPM and CoCo attestation, for secure fleet reporting.
amutable.com
September 22, 2026 at 12:46 PM